eDiscovery Predictive Coding

Home / eDiscovery Services / eDiscovery Predictive Coding

Phases of our eDiscovery Predictive Coding Process

MSI’s eDiscovery predictive coding (or assisted review) services cut cost and time by automating review document prioritization. By following a few simple phases, predictive coding allows for our end-users to isolate the most highly relevant documents for early case assessment in order to understand facts of the case early in the matter.

  • Phase 1
    Key reviewer(s) are giving specific “seed” documents to review and tag as part of the first phase of eDiscovery predictive coding. These “seed” documents are a subset of the overall data that has been ingested into our system.
  • Phase 2
    Reviewer(s) are then given an additional set of documents to refine the eDiscovery predictive coding results. This allows for optimal precision and recall of data within the set(s).
  • Phase 3
    Any conflicts found within the documents tagged in the “seed” set are identified for correction as the eDiscovery predictive coding phase(s) continue.
  • Phase 4
    After both the “seed” and conflict review is 100% complete, each document requiring review is given a responsiveness probability percentage, at which point eDiscovery predictive coding will mark documents as responsive or not responsive, greatly reducing your overall review time.
eDiscovery Predictive Coding
Seed 26%
Precision Review 51%
Conflict Review 76%
Responsive Results 100%

Want More Info?

eDiscovery Data Filtering Now that we've greatly reduced your review set, we'll filter dates, keywords and more!